investigating and addressing publication and other biases in meta-analysis
DESCRIPTION
Studies that indicate a substantial impact of therapy are more likely than other studies to be published, published in English, referenced by other authors, and published numerous times. As a result, such papers are more likely to be found and included in systematic reviews, introducing bias. Another primary source of bias is the low methodological quality of research included in a systematic review. Small studies are more susceptible to all of these biases than large studies. The more significant the treatment impact required for the results to be necessary, the smaller the research. Bias in a systematic review may be detected by looking for a correlation between the size of the treatment effect and the size of the study; such correlations can be analyzed visually and quantitatively. Read More with Us: bit.ly/3fGqLHl Why Statswork? Plagiarism Free | Unlimited Support | Prompt Turnaround Times | Subject Matter Expertise | Experienced Bio-statisticians & Statisticians | Statistics across Methodologies | Wide Range of Tools & Technologies Supports | Tutoring Services | 24/7 Email Support | Recommended by Universities Contact Us: Website: www.statswork.com Email: [email protected] #UnitedKingdom: +44 1618184707 #India: +91 4446313550 WhatsApp: +91 8754467066TRANSCRIPT
Investigating andAddressing Publicationand Other Biases inMeta-analysisAn Academic presentation by Dr. Nancy Agnes, Head, Technical Operations, Statswork Group www.statswork.comEmail: [email protected]
Introduction
Graphical methods for detecting bias
Statistical methods for detecting and
correcting for bias
Conclusion
Outline
TODAY'S DISCUSSION
Studies that indicate a substantial impact of therapyare more likely than other studies to be published,published in English, referenced by other authors,and published numerous times.
As a result, such papers are more likely to be foundand included in systematic reviews, introducing bias.
Another primary source of bias is the lowmethodological quality of research included in asystematic review.
INTRODUCTION
Contd...
Small studies are more susceptible to all of these biases thanlarge studies.
The more significant the treatment impact required for the results tobe necessary, the smaller the research.
Bias in a systematic review may be detected by looking for acorrelation between the size of the treatment effect and the size ofthe study; such correlations can be analysed visually andquantitatively.
Contd...
Funnel plots were initially utilised in educational andpsychological studies.
They're scattered plots of treatment effects calculated fromindividual studies (horizontal axis) vs a metric of study size(vertical axis).
FUNNEL PLOTS
GRAPHICAL METHODS FORDETECTING BIAS
Contd...
Because the precision of calculating the underlying treatment effect improves asthe sample size of research grows, effect estimates from small studies scattermore widely towards the bottom of the graph, decreasing as the sample sizeincreases.
The plot resembles a symmetrical inverted funnel in the absence of bias
Because smaller trials are demonstrating no statistically significant positiveimpact of the therapy (open circles in fig 1 (left) go unreported, reporting biasresults in an asymmetrical funnel plot (fig 1 (centre) with a gap in the bottomcorrect.
Contd...
Fig 1 Hypothetical funnel plots [1]
The combined impact from meta analysisoverestimates the treatment's effect in this case.
Smaller studies are done and analysed withless methodological rigour on average thanmore significant research.
Thus, asymmetry may arise by overestimatingtreatment effects in smaller trials with inferiormethodology quality.
The selection process that decides whether findings are publishedis modelled using "selection models," which are based on thepremise that the study's P value impacts its chance of publication.
The methods may be expanded to estimate treatment effectsadjusted for estimated publication bias, but the lack of strongassumptions about the nature of the selection processnecessitates a large number of studies to cover a wide range of Pvalues.
SELECTION MODELS
STATISTICAL METHODS FOR DETECTINGAND CORRECTING FOR BIAS
Contd...
According to published applications, a meta-analysis of homoeopathic trials andcorrection may explain part of the relationship seen in meta-analyses of theseresearch.
When publication bias is considered, the "correction" of impact estimates iscomplex and a source of continuous dispute.
The modelling assumptions employed may have a significant impact on the results.
Many factors can influence the likelihood of a set of results being published, and it'sdifficult, if not impossible, to predict them all correctly.
Contd...
Furthermore, publication bias is simply one ofthe plausible reasons for treatment-effects-study-size relationships.
As a result, it's best to limit statisticalapproaches to detecting bias rather than fixing itwhen modelling selection mechanisms.
Investigators should attempt to locate allpublished research and seek unpublishedmaterial when conducting a systematic reviewand meta-analysis.
The quality of component studies should bescrutinised as well.
Selection models for publication bias are mostlikely to be useful in sensitivity studies examininga meta-analysis's resilience to potentialpublication bias.
CONCLUSION
Contd...
In most meta-analyses, funnel plots should be utilised to offer a visualassessment of whether treatment effect estimates are related to study size.
Statistical approaches might be used to investigate the evidence for funnel plotasymmetry and alternative reasons for study heterogeneity.
However, these techniques are restricted, especially in meta-analyses based on asmall number of small research.
Contd...
The findings of such meta-analyses shouldalways be taken with a grain of salt.
Combining data from fresh trials statistically witha body of faulty research does not eliminate bias.
When a systematic review shows that theevidence to date is unreliable for one or more ofthe reasons described above, there is presentlyno agreement to guide clinical practice or futureresearch.
